The last meeting with Matthew implanted a really cool idea about solving
equations. Suppose we have an equation f(x) which has finitely many solution
a1, a2, a3, ... an. Then we can say that the solutions of equation f(x) = 0
are
equivalent to the solutions of the equation `(x - a1)*(x - a2) ...
And most of the time I am concerned any version working, so if we worked
from 3.4 backwards that would be great.
On Tuesday, July 1, 2014 10:41:35 AM UTC-5, Ondřej Čertík wrote:
This would be absolutely helpful, as many times I just want to make
sure that Python 3.4 works, etc.
On Tue,
On Wed, Jul 2, 2014 at 8:38 AM, Chris Smith smi...@gmail.com wrote:
And most of the time I am concerned any version working, so if we worked
from 3.4 backwards that would be great.
That's right --- one some difficult PR, I want to first make it work
with Python 3.4, then go backwards down to
I think the solvers should try to abstract two basic ideas:
- Rewriting. You've exposed some ideas here about that. Basically,
rewriting an expression in a way that makes it easier to solve, or to
apply decomposition.
- Decomposition. This means rewriting F(x) as f(g(x)). You then invert
f and g
On Wednesday, July 2, 2014 5:32:56 PM UTC+4, Harsh Gupta wrote:
The ideas is to see equation solving techniques as transformation on
equations
which takes an equation as input and produces a *simpler* equation which
has
the same solutions as the input.
Sounds good so far.
But please
On Wed, Jul 2, 2014 at 1:21 PM, Aaron Meurer asmeu...@gmail.com wrote:
I think the solvers should try to abstract two basic ideas:
- Rewriting. You've exposed some ideas here about that. Basically,
rewriting an expression in a way that makes it easier to solve, or to
apply decomposition.
-
I wrote this up today. In physics.mechanics we often have to sub symbols
for values (or a smaller subset of symbols, i.e. the operating point).
For the huge expressions generated, `subs` is extremely slow. Also, it subs
inside derivatives, which is not ideal (we are currently using a hacky
On Wednesday, July 2, 2014 6:32:56 AM UTC-7, Harsh Gupta wrote:
The last meeting with Matthew implanted a really cool idea about solving
equations. Suppose we have an equation f(x) which has finitely many
solution
a1, a2, a3, ... an. Then we can say that the solutions of equation f(x) =
Does it not work with that? Every value is evaluated.
I think normally, though, you just need to make sure that _eval_evalf
is defined, so that it knows how to evaluate points numerically.
Aaron Meurer
On Tue, Jun 24, 2014 at 4:48 AM, Amit Saha amitsaha...@gmail.com wrote:
Hi all,
I am just